Add gemtext version

content/posts/search-engines-with-own-indexes.gmi

@@ -0,0 +1,199 @@
+DRAFT POST: this post is in progress/under review on envs.net and will be posted on my main site, seirdy.one when it’s ready. Don’t share it yet.
+
+=> gemini://seirdy.one/  seirdy.one
+
+This is a cursory review of all the indexing search engines I have been able to find.
+
+The three dominant English search engines with their own indexes¹ are Google, Bing, and Yandex (GBY). Many alternatives to GBY exist, but almost none of them have their own results; instead, they just source their results from GBY.
+
+With that in mind, I decided to test and catalog all the different indexing search engines I could find. I prioritized breadth over depth, and encourage readers to try the engines out themselves if they’d like more information.
+
+I primarily evaluated English-speaking search engines because that’s my primary language. With some difficulty, I could probably evaluate a Spanish one; however, I wasn’t able to find any Spanish-language engines powered by their own crawlers.
+
+This page is a “living document” that I plan on updating indefinitely. Check for updates once in a while if you find this page interesting. Feel free to send me suggestions, updates, and corrections; I’d especially appreciate help from those who speak languages besides English and can evaluate a non-English indexing search engine. Contact info is in the article footer.
+
+## Methodology
+
+I mainly evaluated link results, and didn’t focus too much on (often glaring) privacy issues, “enhanced” or “instant” results (e.g. Wikipedia sidebars, related searches, StackExchange answers), or other elements.
+
+I compared results for esoteric queries side-by-side; if the first 20 results were (nearly) identical to another engine’s results (though perhaps in a slightly different order), they were likely sourced externally and not from an independent index.
+
+I tried to pick queries that should have a good number of results and show variance between search engines. An incomplete selection of queries I tested:
+
+* “vim”, “emacs”, “neovim”, and “nvimrc”: Search engines with relevant results for “nvimrc” typically have a big index. Finding relevant results for the text editors “vim” and “emacs” instead of other topics that share the name is a challenging task.
+* “vim cleaner”: should return results related to a line of cleaning products rather than the Correct Text Editor.
+* “Seirdy”: My site is relatively low-traffic, but my nickname is pretty unique and visible on several of the highest-traffic sites out there.
+* “Project London”: a small movie made with volunteers and FLOSS without much advertising. If links related to the movie show up, the engine’s really good.
+* “oppenheimer”: a name that could refer to many things. Without context, it should refer to the physicist who worked on the atomic bomb in Los Alamos. Other historical queries: “magna carta” (intermediate), “the prince” (very hard).
+
+## General indexing search-engines
+
+### Large indexes, good results
+
+These are large engines that pass all the above tests and more.
+
+1. Google: the biggest index. Allows submitting pages and sitemaps for crawling, but requires login. Powers a few other engines:
+
+* Startpage
+* Runaroo
+* SAPO (Portuguese interface, can work with English results)
+
+2. Bing: the runner-up. Allows submitting pages and sitemaps for crawling, but requires login. Its index powers many other engines:
+
+* Yahoo
+* DuckDuckGo²
+* AOL
+* Qwant³
+* Ecosia
+* Ekoru
+* Privado
+* Findx
+* Disconnect Search⁴
+* PrivacyWall
+* Lilo
+* SearchScene
+* Peekier
+* Oscobo
+* Million Short
+* Yippy search⁵
+* Lycos
+* Givero
+* Swisscows
+* Ask.moe⁶
+* Partially powers MetaGer by default; this can be turned off
+* At this point, I stopped adding Bing-based search engines. There are just too many.
+
+3. Yandex: originally a Russian search engine, it now has an English version. Some Russian results bleed into its English site. Allows submitting pages and sitemaps for crawling, but requires login. Powers:
+
+* Epic Search
+* Occasionally powers DuckDuckGo’s link results instead of Bing.
+
+4. Mojeek: Claims to be privacy-oriented. Quality isn’t at Google/Bing/Yandex’s level, but it’s not bad either. If I had to use Mojeek as my default general search engine, I’d live.
+5. Petal search: gopetal.com and petalsearch.com. A very new engine developed by Huawei. Surprisingly good results; it passed all the listed tests. Requires an account to submit sites. I discovered this via my access logs.
+
+=> https://www.gopetal.com/  gopetal.com
+=> https://petalsearch.com/  petalsearch.com
+
+### Smaller indexes, relevant results
+
+These engines pass most of the tests listed in the “methodology” section.
+
+* Right Dao : very fast, good results. Passes the tests fairly well.
+* Gigablast : It’s been around for a while and also sports a classic web directory. Searches are a bit slow, and it charges to submit sites for crawling. It powers Private.sh. Gigablast is tied with Right Dao for quality.
+* Gowiki : Very young, small index, but shows promise. I discovered this in the seirdy.one access logs.
+
+=> https://rightdao.com  Right Dao
+=> http://gigablast.com/  Gigablast
+=> https://private.sh  Private.sh
+=> https://gowiki.com  Gowiki
+
+### Smaller indexes, hit-and-miss
+
+These engines fail badly at a few important tests.
+
+* seekport : The interface is in German but it supports searching in English just fine. The default language is selected by your locale. It’s really good considering its small index; it hasn’t heard of less common terms (e.g. “Seirdy”), but it’s able to find relevant results in other tests.
+* Exalead : slow, quality is hit-and-miss. Its indexer claims to crawl the DMOZ directory, which has since shut down and been replaced by the Curlie directory. No relevant results for “Oppenheimer” and some other history-related queries. Allows submitting individual URLs for indexing, but requires solving a Google reCAPTCHA and entering an email address.
+* wbsrch : In addition to its generalist search, it also has many other utilities related to domain name statistics. Failed multiple tests. Its index is a bit dated; it has an old backlog of sites it hasn’t finished indexing.
+* ExactSeek : small index, disproportionately dominated by big sites. Failed multiple tests. Allows submitting individual URLs for crawling, but requires entering an email address and receiving a newsletter. Webmaster tools seem to heavily push for paid SEO options.
+
+=> http://www.seekport.com/  seekport
+=> http://www.exalead.com/search/  Exalead
+=> https://curlie.org  Curlie
+=> https://wbsrch.com/  wbsrch
+=> https://www.exactseek.com/  ExactSeek
+
+### Unusable engines, irrelevant results
+
+Results from these search engines don’t seem at all useful.
+
+* YaCy: community-made index; slow. Results are awful/irrelevant, but can be useful for intranet or custom search.
+* Scopia: only seems to be available via the MetaGer metasearch engine after turning off Bing and news results. Tiny index, very low-quality.
+* Active Search Results : very poor quality
+* Crawlson: young, slow. In this category because its index has a cap of 10 urls per domain. I initially discovered Crawlson in the seirdy.one access logs. The site seems to be down right now, so I didn’t link it.
+
+=> https://metager.org  MetaGer
+=> https://www.activesearchresults.com  Active Search Results
+
+## Non-generalist search
+
+These indexing search engines don’t have a Google-like “ask me anything” endgame; they’re trying to do something different.
+
+* Wiby: I love this one. It focuses on smaller independent sites that capture the spirit of the “early” web. It’s more focused on “discovering” new interesting pages that match a set of keywords than finding a specific resources. I like to think of Wiby as an engine for surfing, not searching. Runaroo occasionally features a hit from Wiby. If you have a small site or blog that isn’t very “commercial”, consider submitting it to the index.
+* Quor: seems to mainly index large news sites.
+
+=> https://wiby.me  wiby.me
+=> https://wiby.org  wiby.org
+=> https://www.quor.com  Quor
+
+## Other languages
+
+I’m unable to evaluate these engines properly since I don’t speak the necessary languages. English searches on these are a hit-or-miss. I might have made a few mistakes in this category.
+
+### Big indexes
+
+* Baidu: Chinese
+* Qihoo 360: Chinese. I’m not sure how independent this one is.
+* Sogou: Chinese
+* Yisou: Chinese
+* Naver: Korean.
+* Seznam: Czech, seems relatively privacy-friendly. Discovered in the seirdy.one access logs.
+* Cốc Cốc: Vietnamese
+
+=> https://search.naver.com  Naver
+=> https://www.seznam.cz/  Seznam
+=> https://coccoc.com/search  Cốc Cốc
+
+### Smaller indexes
+
+* Parsijoo: Persian
+* search.ch: Regional search engine for Switzerland; users can restrict searches to their local regions.
+* fastbot: German
+* Moose.at: German (Austria-based)
+
+=> http://www.parsijoo.ir/  Parsijoo
+=> https://search.ch  search.ch
+=> https://www.fastbot.de/  fastbot
+=> https://www.moose.at  Moose.at
+
+## Misc
+
+* Ask.com: the main site shut down, but subdomains like uk.ask.com are still alive. They claim to outsource search results. The results seem similar to Google, Bing, and Yandex; however, I can’t pinpoint exactly where their results are coming from.
+* Not evaluated: Apple’s search. It’s only accessible through a search widget in iOS and macOS and shows very few results. This might change; see the next section.
+* Partially evaluated: Infinity Search young, small index. It recently split into a paid offering with the main index and Infinity Decentralized, the latter of which allows users to select from community-hosted crawlers. I managed to try it out before it became a paid offering, and it seemed decent; however, I wasn’t able to run the tests listed in the “Methodology” section. Allows submitting URLs and sitemaps into a text box, no other work required.
+
+=> https://uk.ask.com  uk.ask.com
+=> https://infinitysearch.co  Infinity Search
+=> https://infinitydecentralized.com/  Infinity Decentralized
+
+## Upcoming engines
+
+These engines aren’t ready yet; their indexes are either in a proof-of-concept phase with a handful of sites, or aren’t available yet.
+
+* Apple: given the activity of the AppleBot crawler lately, their index will almost certainly grow to a size large enough to power a general search engine soon. Check your server’s access logs; there’s a good chance it’s crawled your site if you have a few backlinks.
+* Brave announced that it will start its own independent search engine based on the now-discontinued Cliqz. I don’t endorse the company, but I hope its results will be available through metasearch engines like Searx.
+
+=> https://brave.com/search/  Brave Search
+
+## Acknowledgements
+
+Some of this content came from the Search Engine Map and Search Engine Party. A few web directories also proved useful.
+
+=> https://www.searchenginemap.com/  Search Engine Map
+=> https://searchengine.party/  Search Engine Party
+
+
+## Notes
+
+¹ Yes, “indexes” is an acceptable plural form of the word “index”. The word “indices” sounds weird to me outside a math class.
+
+² DuckDuckGo has a crawler called DuckDuckBot. This crawler doesn’t impact the linked results displayed; it just grabs favicons and scrapes data for a few instant answers
+
+³ Qwant claims to also use its own crawler for results, but it’s still mostly Bing. Try a side-by-side comparison; I found that it doesn’t seem to have anything besides Bing results.
+
+⁴ Disconnect Search allows users to have results proxied from Bing or Yahoo, but Yahoo sources its results from Bing.
+
+⁵ Yippy claims to be powered by a certain IBM brand (a brand that could correspond to any number of products) and annotates results with the phrase “Yippy Index”, but a side-by-side comparison with Bing and other Bing-based engines revealed results to be nearly identical.
+
+⁶ Ask.moe was working on a FLOSS indexer; its search page stated an intention to switch to it from Bing at one point. This statement has since been removed.
+
+=> https://git.sr.ht/~danskeren/spider.moe  FLOSS indexer