A New Search Engine That Doesn’t Rank Sites Based on Their Authority | Hacker News Reader

A New Search Engine That Doesn’t Rank Sites Based on Their Authority

ortingo.com

32 points_t9gb42 comments

I just randomly searched for the first thing that popped into my head: “tallest man in the world”

The entire page of results were for an SEO marketing book. Nothing even remotely related to what I searched for.

What is this good for?

I searched for C++ SIMD library and it was PHP links. So it’s got a lot of work needed.

I searched for a couple of things such as my name, my company and found nothing where on Google my name brings up 138k results. Now Google goes overboard but going from that to 0... Probably wrong.

IMO it seems like you're trying to do too much at once. If these are just random throw away projects, that's fine.

But if you're actually trying to create useful services, creating a bunch of half-baked services and moving onto the next one is a recipe to make nobody happy.

Some (hopefully) constructive feedback:

* The site is not doing you any favors in representing your web-design agency. Personally I'm not a fan of the aesthetic, but that's neither here nor there. The big issue is the usability of the site. I'd spend some time learning and practicing proper UX.

* There doesn't seem to be any mention of what the "Ortingo Plans" are for. What is this storage for?

* The marketing offer is a) suspect and b) underwhelming. "Virtually guaranteeing" a view count makes me think of those old sketchy "traffic blaster" sites. And 750 views over 30 days is basically nothing, especially for $500 when those views are completely untargeted. Views themselves are worthless. The reason people pay for views (via ads) is to get some percentage of those viewers to convert into buyers.

* I think enough has been said on this thread about the search results so I don't need to add anything there. But the UX is not good on the search results. I'd look up Jakob's law if you aren't familiar with it.

The animations are repulsive. Every back button press results in vertigo. Ortingo storage, ortingo web design agency, ortingo search? Some bootcamp graduate has some major Messiah complexes. This thing is built like an air mattress.. pins included.

The results..well the other comments speak enough about how empty the entire site is. What'd you do, scrape a megabyte of the internet and call it a day?

So I tried my first search: "proof of correctness".

Results:

    Promoted post: Losing 29 GB of Data:

    You don't have to go through what I went through, unless you already did,
    and if you did, I hope... by Daniel Dahdal (Founder of Ortingo)

Post not listed as an ad:

    Everything I Know To Be True About Getting $5,000+ Website Clients — My Billie Designs

Second search: "Trump".

Results: 5 different sites with the same text:

    Login

    https://asp.schoolmessenger.com/nwsd203/loginpicture.img.php

    President Donald Trump has vowed to bring more manufacturing jobs
    to America, so here are 50 manufac...

(Plus the same promoted post as above.)

So, total fail on a moderately obscure subject, and total fail on a popular keyword.

Is this a joke site?

Any search engine except google itself is a joke site at this point, there's no way for any company to reach that level of collected data and subsequently extracted heuristics. They have a huge head start and will just collect more in the time a new company will take to get to the same level.

Sure there is. You can use common crawl[0] to start with. You don't need to be the 'next Google', you just need to give relevant results and that's a completely solved issue at this point.

[0]https://commoncrawl.org

Eh, there's a difference between giving okay results and the ones you need in every possible case.

Kind of how it's easy to make a robot that works 80% of the time for some task but the 99.999% of the time is the challenging bit. Which effectively means you have to go open google when the thing won't give you what you need after a few failed searches and some wasted time.

Like Yahoo and Bing may be alright, with their aggregators like DDG giving even slightly better results, but it still falls short of the big G. You can be good, but you will never be better.

Just turned the "search engine" into a discussion platform. Decided a good way to have the platform populated with data, was by turning it into that, rather than the search engine it once was, which was a horrendous failure, due to the weak crawler, which was responsible for the bad results you all saw in your experience with the platform. I kept the search functionality, but as you may have already figured, it doesn't search for web results, rather discussions posted by users on the platform.

I don't think this search engine is quite ready for its prime time. Almost all the results are from graph subdomain of facebook. My guess is that maybe the engine hasn't crawled enough of anything else and is still stuck at the large sites?

More unusual is that the results don't seem to be relevant to the search query as has been mentioned by others. I have experimented with a very simple search engine before based Sonic index with post-match sorting based on basic query word distance rules and even that worked much better, although I most likely had crawled much less data into the database.

It feels that it would be more interesting to have a search engine that by default excludes facebook, linkedin, twitter and other social networks. Rank the large ones out, keep the little blogs straight from the source. Unfortunately I suspect this is not too attractive of an idea for most companies.

Search for "hacker news" https://ortingo.com/?q=hacker+news&page=1&search_type=web

results in:

"gatx tank car close up" "20 Products With Giant Markups"

Yeah, i think the authority method is better to be honest.

Is it a joke? Seriously, each request give me a bunch of random results not related at all with the request.

Seems like it... Searching for "google" brings back a load links to images (that might be hosted on some google static url perhaps), with some random summary text that seems to have nothing to do with the image...

Does anybody know where to find information on what the "consistently fair algorithm to find relevant information" which "doesn't rank websites based on their authority" actually is? Or any other technical info, like whether they build their own index or (like pretty much every search startup) simply buy their results from Bing?

Also note that it is ad funded - "we virtually guarantee your post to receive at least 750 views per 30-day period (around 25 views daily), at $500 (around $16 daily)", which (IMHO) sets up an immediate potential conflict with user needs.

Weirdo unnecessary animation as the almost nonexistent results grow in makes the site look and feel janky. Individual results with relatively small text in bubbles is MUCH harder to scan than traditional vertical results scanned top to bottom.

Since there isn't really much in the way of results I'm charitably assuming the backend isn't meaningfully finished yet and this is just meant to showcase not the results thus far so much as the UI which unfortunately because of the above problems is uniformly terrible.

Lose the animations nobody cares and please consider a more traditional listing of results.

https://ortingo.com/?q=cats&page=1&search_type=image

I get zero results, this project is doomed.

I searched for 'ColdFusion language' and all the entries concerned PHP and Perl, nice job. Guess I should have searched for PHP and Perl and then I might have gotten ColdFusion ;<).

Results arent good unfortunately but for those liking the idea but interested in something similar, check out metaphor.so - results are better than Google in my view.

Hey all, I just looked at your comments! As critical as these comments were, it was obvious constructive feedback was the intention, which I really appreciate. Good time to realize it's time for a massive re-engineering and redesign of this miserable failure of a project, quite frankly.

I clicked on the submitter's username. This is his personal site. Which used to be a blog. Which isn't anymore.

This is basically a Show HN.

So, I have questions.

What do you mean that it doesn't rank sites based on authority? What do you mean by authority? Because PageRank (for example) ranks sites based on relevance.

What are you ranking sites based on?

I hope it's a research project and not an actual search engine. It's not bad, it's random in every field of response.

Searched for opensearch ssl setup, second result:

- Title is a chapter from mysql documentation;

- Site link is a burger restaurant in US;

- Description is "Setup instructions for Mail-in-a-Box";

So what happens when there are too many "authorities" in their space: how do you decide which one to show first and which one to show next?

Ralistically, is anyone really going to go past page 2 of results or grow frustrated and go back to Google/DDG?

Interesting concept, but man the results are bad. Definitely not a “new standard.”

"Rocket League" gets 2 results "How We Acquired 100K Early Bird Signups with Zero Marketing Budget" and "Key Ingredients To Make Your App Go Viral". Same results if put in quotes.

Works Great: My first search resulted in this:

Connection failed: User 'u126801970_ortingo' has exceeded the 'max_user_connections' resource (current value: 15)

Interesting idea but the search matches aren't great.

Also the UI is not easy to scan, the list view of search results has been standardized for decades now for a reason.

Oops

> Connection failed: User 'u126801970_ortingo' has exceeded the 'max_user_connections' resource (current value: 15)

I tried to search, but was thwarted by the UI. I wasn't able to get more out of it than the logo and the cookie notice.

duckduckgo has been marketed to death and still has only 2% marketshare. It's one thing to code a search engine ,entirely another to make one peple will use.

DuckDuckGo is often a superior search engine to alternatives, certainly less human bias, but their name does not help them. Who says "DuckDuckGo it?" They need to shorten the name considerably, either to Duck, or pick a different one entirely.

Who is going to say "Duck it" though?

Duckduckgo isn't a 'search engine', they just get their results from Bing, they don't crawl shit themselves. They're good at marketing and convincing technically illiterate people to use them, that's all.

Look at a (non-American) search engine like Quant[0] if you want to see something people actually use.

[0]https://www.qwant.com/?l=en

Qwant also uses Bing.

Did anybody ever get even a single result for any query?

This was useless - the results were trash. Sorry.

> everyone gets a chance to become an authority

Oh, great.

Unable to find any image results for cats.

or doesn't rank them at all....

I just searched: Laravel

No results... definitely needs work... lots of work.

It's terrible.

Very interesting because of the weird results(albeit few) from random websites. It's a great discovery tool.

Now hoping for search engine that favors text-heavy sites and punishes paywalls

A search for "dog" yeilded zero results, terrible

Reply on news.ycombinator.com