#yacy — Public Fediverse posts

@newsgroup Does this project connect to the existing Yacy network, or to another separate network of peers?

#Yacy

Do you know how I can integrate the #peertube #searchengine https://sepiasearch.org/ in the results of a #yacy instance ?

#Froscon ; Vorträge. die ich besucht habe:
Wahrscheinlich kennt kaum jemand die Suchmascine #Yacy. Hab 2014 damit rumgespielt, installiert auf einem Bananapi mit echtem SATA und Platte und war begeistert
https://www.wormser-region.de/index67e2.html

Gestern war ich bei #Froscon in einem Vortrag und jetzt gibts Yacy mit KI und auch jetzt wieder wird das Programm komplett lokal installiert 👍 👍 👍 😃
Video: https://media.ccc.de/v/froscon2026-3643-ki-chat_mit_eigener_suchmaschine_rag_in_yacy

Hat schon mal irgendwer #YaCy auf #YunoHost zum laufen gebracht? Ich kann die Adminseite nicht aufrufen. 😭

@BrideOfLinux
Let's start completing the list of independent search engines

#mwmbl
#yacy
#PriEco

Update: I just read a bit more about #NeoSearch. Quote "[...] using Google's search index for now.". This means it is not like the engines mentioned above.

@hyperreal
You put some effort into this thoughtful reply.

The #Firefox browser I use does have built-in translation, which I sometimes use, but the results can be poor. Other AI produces better quality text.

I am working on a script that searches the internet in shell with #Yacy and no browser. The text is translated automatically and added to the same page, which I can read with no clicks.

#AI can do more than #translate. It can also summarize the web pages found and answer questions.

@domdel @Framasoft #yacy sur #android ??!! je n’ai jamais essayé, mais sur PC ça marche bien.

Par contre si on l’utilise aussi comme crawler, ça pompe pas mal d’energie et de Go.
Et les résultats, je trouve, ne sont pas souvent à la hauteur, et l’ordre par default est plus proche du random qu’autre chose.

Ceci dit je recommande à tout les bidouilleurs qui ne sont pas allergique aux LLM et autres 'agent'.

@Framasoft On peut aussi parler de #YaCy comme alternative sous #android https://yacy.net/installation/android/

J'utilise le moteur de recherches #YaCy . Il est entièrement configurable. Ça paraît fastidieux mais une fois configuré aux petits oignons, on obtient des résultats fiables, sans pubs sans bots etc.

@cquest @Framasoft Avec #YaCy c'est peut être configurable; je regarde.
https://yacy.net/download_installation/

@mmeier
You can lower the speed of crawling down to as low as 10 pages per minute if you like.

What I like about #Yacy is it produces uncensored search results.

#censorship

YaGo: как я хотел бесплатный self-hosted Tavily API, а в итоге воскресил YaCy

Всё началось не с мечты про «поисковик нового поколения». Мне понадобился быстрый self-hosted Tavily-compatible API для собственных рабочих и личных AI-решений: без оплаты за каждый запрос, без внешнего сервиса в обязательной цепочке и с индексом, содержимое которого контролирую я сам. Тут я вспомнил про YaCy. Когда-то я уже поднимал его ноду. Идея мне нравилась, а реализация — заметно меньше: Java, тяжёлая машина и примерно шесть секунд ожидания ответа на моей тогдашней установке. Для человека, который один раз нажал Enter, это ещё можно пережить. Для агента, делающего несколько поисков, уточнений и extract подряд, это превращает один шаг в минутный перекур. Поэтому вместо ещё одной обёртки над чужим поиском я оставил от YaCy сетевой протокол и начал собирать поисковую ноду заново: на Go, с отдельным краулером, embedded storage, нормальным API и ranking pipeline из современных работ по information retrieval. Под катом — немного сетевой археологии, Bleve, bbolt, gRPC, BM25, LambdaMART и рассказ о том, как задача «дайте локальный endpoint для AI-агентов» постепенно превратилась в реинкарнацию YaCy.

https://habr.com/ru/articles/1060298/

#yacy #поиск #selfhosted #ai #llm #tavily #яндекс #google

@MrClicko @orbiterlab

YACY habe ich noch nicht zum Laufen gebracht auf einer IPv6 only Maschine.

Die Seeds scheinen nur auf IPv4 Host vorhanden zu sein.

https://yacy.net/

#yacy #ipv6

@meirisoda
I found you! How? I was crawling the web with #Yacy, noticed a link pop up in the crawl results, with a title of "soda is online". It was your website: https://meirisoda.online/ ~ I wondered, what does it mean that soda is online? I clicked the link and saw you are also online here at Mastodon.

And I like your profile image.... reminds me of the star of Bethlehem.

@athena_pronaia @arteesetica Salvo que montes tu propia instancia con #YaCy https://en.wikipedia.org/wiki/YaCy

@mstadmin
#Yacy allows searching any #Mastodon post from any server.... IF it has crawled it. I have my server set up to crawl certain feeds on a schedule, so the search results will be up-to-date. And Yacy is not for Mastodon only, but for any web page that has been crawled.

Regarding #Elasticsearch, is that a search option available on your Mastodon instance but requires you to be logged in to use it? You can search only your posts, and not those from other accounts, right?

@mstadmin
Posting online is like an open diary, or a log of your life, if you want it to be. And yes, it is helpful to have it documented so you can come back later to refresh your memory.

#Elasticsearch is not something I know about, but that is good if it works for you.

I use #Yacy when I need to search my own posts. Example: Yacy finds the last time I mentioned #Mary in a #Mastodon post....

how many hidden monitored buzzword tags are there for heaven's sake? I approve, it should help unify certain posts #json #cvs #gitlab #agile #django unchained #ci/cd #pipeline #java #yacy #jar #postgres #software #devops

@orbiterlab It's a pitty that the only description/how-to of all these shiny new #yacy #AI features is included in German-spoken video only. Are you planning to make at least brief docs page? Or English talk or anything?

New blog post: https://blog.mei-home.net/posts/yacy-part-2-crawling-backburner/

The second post in the YaCy series, talking about crawling the web to fill the search inxed, and where I see issues.

Also putting the project on the backburner for now, because...well, it's really gotten to the size of a main project, and I already have one of those at the moment.

#HomeLab #Blog #YaCy

Hmmm, really confused about YaCy's crawling code at the moment. It looks to me like there's no handling for 429 HTTP codes. The URL which returned that code is just dropped, same as actual errors like 404. I would say the right approach would be to reduce the crawl frequency for the site and try again later on?

Or am I overlooking some Netiquette here?

#HomeLab #web #YaCy

Time to dig into YaCy's crawler code it seems. I'm getting 429 codes a lot from pkg.go.dev. And it looks like YaCy's crawler in those cases just skips the URL, instead of just waiting a little bit and reducing the crawl frequency.

I'm starting to think that this entire web crawling business is not, in fact, easy.

#HomeLab #YaCy

Hmpf. My first dive into web crawling was already pretty disappointing. It seems that you can't crawl a GitHub project's issues, because the /issues page doesn't actually contain any links to any issues.

Same for PRs.

What's even worse: When you open an issue's page, the HTML doesn't even contain the content of the description or any of the issue comments.

So looks like if I want to crawl GH for my YaCy search index, I will have to write a bespoke crawler.

#HomeLab #YaCy #GitHub

New blog post: https://blog.mei-home.net/posts/yacy-part-1-deployment/

This is the first post in what's very likely going to be a whole series about the YaCy distributed P2P search engine.

In this initial post I describe YaCy a bit and go through the problems I've seen while deploying it into my k8s cluster.

#Blog #YaCy #HomeLab

et hop un petit noeud Yacy pour ajouter à l'outil de recherche P2P sur
#yunohost et mon petit serveur auto-hébergé dans ma cuisine

parce que j'ai fermé la porte à tous les crawlers ...!!!
sauf Qwantbot , pour le moment.

je retrouve facilement mes mots balises maintenant , genre "partageoire"
et peut remonter à la surface des vieux sites oubliés.

Yacy ne me mange pas trop de ressource
👍

#yacy #autohebergement

Oh no. YaCy got OOM-killed in the middle of that crawl, and now the crawl queue is gone. And a re-crawl won't help directly, because that will not crawl pages which it has already seen. Do I'd need to delete everything related to the page from the index and begin a fresh crawl.

#HomeLab #YaCy

I really need to start to think harder about my crawler configs. I just decided to launch a crawl on my birth town's website. Because I previously found that YaCy's "crawl only start domain" config doesn't include subdomains, I decided to set my own regex.

And I did not think hard enough, because I used `.*<city>.de/.*`. Which of course also catches stuff like <org>-in-<city>. And there seem to be a lot of those. 😅

#HomeLab #YaCy

@fanta Si el índice local no ocupa mucho me parece una solución. Supongo que sin activar el índice distribuido ocuparía mucho más, a cambio de más privacidad.

Lástima que esté escrito en #Java.

#YaCy