? Editing: Post:21.body Save Delete Cancel
Initial sync in progress...

List of ZeroNet sites

Follow in NewsfeedFollowing
+ Add new site
Loading...
stickied

Title

Body
^1 ^2 added ━ submitted by user_name
More sites

 

Follow in NewsfeedFollowing

Search Engine - Horizon

^18 ^19 blurhy posted on Aug 23, 2018
Please sign innew comment
Sign in as...
Submit comment
You are running out of your allowed space, please contact the site's admin at unknown to raise your limit.
user_nameadded ^1 ^2
Reply
Body
snifflzon Feb 21, 2019 ^1 ^2
Reply

looks like the project is shut down on github

anonymeon Jan 27, 2019 ^1 ^2
Reply

Address not working

blurhyon Sep 23, 2018 ^1 ^2
Reply

Address changed

quantum2on Sep 17, 2018 ^1 ^2
Reply

+1

blurhyon Sep 12, 2018 ^1 ^2
Reply

The code is ok.So if there is any bug.Please tell me

lithiunon Sep 12, 2018 ^1 ^2
Reply

Thx

remixon Sep 05, 2018 ^1 ^2
Reply

ty guys

mike32on Sep 03, 2018 ^1 ^2
Reply

The best project!

blurhyon Sep 01, 2018 ^1 ^2
Reply
blurhyon Sep 01, 2018 ^2 ^3
Reply

I'm rafactoring my code.So don't use Horizon this few weeks.

blurhyon Sep 01, 2018 ^1 ^2
Reply

Please issue on github or PR https://github.com/blurHY/Horizon

deadwebon Sep 01, 2018 ^1 ^2
Reply

Nice!

ssdifnskdjfnsdjkon Aug 30, 2018 ^1 ^2
Reply

blurhy

Hi, currently the search is not working, see: https://snag.gy/nB2FHT.jpg (Windows & Firefox)
Site was refreshed (F5) and updated by Update button.
Happens in Firefox on Windows & Ubuntu Linux.

r8seon Aug 28, 2018 ^1 ^2
Reply

time to load the page and everything goes white. =/

blurhyon Aug 26, 2018 ^3 ^4
Reply

My Final Idea:

The crawling system needs a Main Crawler

  • Main Crawler publish the need-crawl sites list.(Actully, it is that some crawler wants to crawl then it send a request (like posting a comment) to the sitelist board via zeronet and the crawler gets an id.Then the main crawler publish the zite with new content that describes each crawler's sites list (to crawl) )
  • Each crawler only crawls the specified zites.Then publish the crawled data.(keywords,phrases,list of links that links to new zites)
  • Only main crawler discovers new zites.

@ssdifnskdjfnsdjk ,@daniell

I will write a blog post with a svg

blurhyon Aug 26, 2018 ^1 ^2
Reply

Distributed crawling needs a main peer.Or there will be duplicate links in database.
But we can make a zeronet plugin,that generates the keywords for each sites that doesn't appear in public database then publish the keywords data to my search engine.. @ssdifnskdjfnsdjk

ssdifnskdjfnsdjkon Aug 26, 2018 ^1 ^2
Reply

blurhy: @ssdifnskdjfnsdjk Do you mean distributed crawling.

yes, relying to one person to run some computer in order to supply new content for the SE (search engine) is not ideal, you probably know it (single point of failure)

Better if more people can contribute their computers to crawling (decentralization). Someone mentioned if making a zeronet plugin would not be solution.
OR distributed way of crawling (i am not a programmer to know if you already do it or if that will be possible)

blurhyon Aug 26, 2018 ^2 ^3
Reply

and i uploaded the spider to github https://github.com/blurHY/ZeronetSpider

blurhyon Aug 26, 2018 ^1 ^2
Reply

@ssdifnskdjfnsdjk Do you mean distributed crawling.

blurhyon Aug 26, 2018 ^1 ^2
Reply

ssdifnskdjfnsdjk: @blurhy if you are able to figure out a decentralized way of crawling, that would be better than relying on single person computer or server running
@daniell @blurhy

Unsure at which database size the Zeronet start lagging or when search will be too slow. Currently i seen 3 seconds delay, in results being shown, so far not problem but in future?

Because of the crawler crawls keywords.So you should not use sentence as search query.Or it will be slow.

ssdifnskdjfnsdjkon Aug 26, 2018 ^1 ^2
Reply

daniell: > If you want, I can host this spider in the host I'm buying.The creator of 0net proxy also can do it.

@blurhy if you are able to figure out a decentralized way of crawling, that would be better than relying on single person computer or server running

@daniell @blurhy

Use optional files and merger zites.

Unsure at which database size the Zeronet start lagging or when search will be too slow. Currently i seen 3 seconds delay, in results being shown, so far not problem but in future?

blurhyon Aug 26, 2018 ^1 ^2
Reply

The github repo https://github.com/blurHY/Horizon ,i uploaded just now @daniell @ssdifnskdjfnsdjk

daniellon Aug 25, 2018 ^1 ^2
Reply

blurhy: The right side container is for zerowiki

You can also support Kiwipedia, and filter duplications.
For sample, if ZeroWiki article is already in Kiwipedia, prioritize Kiwipedia.

daniellon Aug 25, 2018 ^1 ^2
Reply

blurhy: I added some Cross Origin Search features.That's useful.Added zerosites,zerotalk,plan to add zerowiki and more

Would be nice if you add ZeroWiki cards.

daniellon Aug 25, 2018 ^1 ^2
Reply

blurhy: i added a lot of features today,but not publish yet.i will add that tomorrow

Do you have a GitHub repo for issue tracking and pull requests?

daniellon Aug 25, 2018 ^1 ^2
Reply

blurhy:

Also I'm going to write a spider that crawls user data and get the keywords.That spider will returns topic url.

If you want, I can host this spider in the host I'm buying.
The creator of 0net proxy also can do it.

daniellon Aug 25, 2018 ^1 ^2
Reply

ssdifnskdjfnsdjk:

Another thing is, when i search for "nettalk", it shows two entries both for same site, one .bit one standard address,

Zirch shows the two addresses (Bitcoin and .bit) in the same result.

daniellon Aug 25, 2018 ^1 ^2
Reply

blurhy: The data maybe huge.

Use optional files and merger zites.

daniellon Aug 25, 2018 ^1 ^2
Reply

ssdifnskdjfnsdjk: If you find links on the sites it is linking to, i think you may discover maybe more than double number of sites For example GFWTalk is large forum that can contain many external links.. In the sidebar of the Important Zites the @krixano made a search in ZeroUp and ZeroMe also.

Good idea, to crawl zite's links like Google does.
And show ZeroMe/Forums posts, ZeroUp files, KopyKate videos, etc.

daniellon Aug 25, 2018 ^1 ^2
Reply

blurhy: The data of 0list and others is not enough for the search engine,i think

Use a button in every result, then user can suggest improvements, and you/zite owner approves.
Also, a button where zite owner can claim ownership.

blurhyon Aug 25, 2018 ^2 ^3
Reply

The right side container is for zerowiki

blurhyon Aug 25, 2018 ^3 ^4
Reply

I added some Cross Origin Search features.That's useful.Added zerosites,zerotalk,plan to add zerowiki and more

blurhyon Aug 25, 2018 ^3 ^4
Reply

ssdifnskdjfnsdjk: @blurhyWhen you add this search query: create zeronet site -zerosecurityyou see alot of irrelevant results, it can be handy to allow user use the operators like exclusion (-zerosecurity)Next thing that would help user filter the results is to allow requiring exact phrase which have to exist, example query:+"create site" OR "make a website" +zeronet +tutorial -zerosecurity
(same syntax that most popular search engine - Google is using)This would greatly help user to filter out exactly what is needed.

i added a lot of features today,but not publish yet.i will add that tomorrow

ssdifnskdjfnsdjkon Aug 25, 2018 ^2 ^3
Reply

@blurhy
When you add this search query: create zeronet site -zerosecurity
you see alot of irrelevant results, it can be handy to allow user use the operators like exclusion (-zerosecurity)
Next thing that would help user filter the results is to allow requiring exact phrase which have to exist, example query:
+"create site" OR "make a website" +zeronet +tutorial -zerosecurity

(same syntax that most popular search engine - Google is using)
This would greatly help user to filter out exactly what is needed.

blurhyon Aug 24, 2018 ^3 ^4
Reply

ssdifnskdjfnsdjk: 1 . > blurhy: The data maybe huge.
is it a problem? What about using merger sites to distribute data across multipe sites/DBs?
2 . Another thing is, when i search for "nettalk", it shows two entries both for same site, one .bit one standard address, may be handy to keep only one link (i prefer standard address as .bit can expire).
3 . Another think is that nettalk search query returned a topic URL which is nice, but apparently it does not find topics from ZeroTalk? Can you add ZeroTalk?
4 . Next think, the URLs are light green color, to better see it, it can be in darker green so there is bigger contrast

I'm making a filter feature.That can be used for filter. For example, same title / same domain / only sites(root url of sites)
Also I'm going to write a spider that crawls user data and get the keywords.That spider will returns topic url.

ssdifnskdjfnsdjkon Aug 24, 2018 ^2 ^3
Reply

1 . > blurhy: The data maybe huge.

is it a problem? What about using merger sites to distribute data across multipe sites/DBs?

2 . Another thing is, when i search for "nettalk", it shows two entries both for same site, one .bit one standard address, may be handy to keep only one link (i prefer standard address as .bit can expire).

3 . Another think is that nettalk search query returned a topic URL which is nice, but apparently it does not find topics from ZeroTalk? Can you add ZeroTalk?

4 . Next think, the URLs are light green color, to better see it, it can be in darker green so there is bigger contrast

blurhyon Aug 24, 2018 ^2 ^3
Reply

The data maybe huge.

ssdifnskdjfnsdjkon Aug 23, 2018 ^2 ^3
Reply

blurhy: The data of 0list and others is not enough for the search engine,i think

If you find links on the sites it is linking to, i think you may discover maybe more than double number of sites For example GFWTalk is large forum that can contain many external links.. In the sidebar of the Important Zites the @krixano made a search in ZeroUp and ZeroMe also.

blurhyon Aug 23, 2018 ^2 ^3
Reply

The data of 0list and others is not enough for the search engine,i think

blurhyon Aug 23, 2018 ^2 ^3
Reply

crawl links from zerosites\0list\new zeronet sites and visit the target zites of the links.Then use nlp library analyse the content of the zites to get the keywords.

daniellon Aug 23, 2018 ^1 ^2
Reply

Take a look at Zirch Automated Search's source code.
It reads data from 0list.
Horizon can get zites from 0list, ZeroSites and others.

blurhyon Aug 23, 2018 ^3 ^4
Reply

Crawler1:Selenium+Chrome+Mysql to scrape webpage.
Crawler2:Directly scrape the zites' database

This page is a snapshot of ZeroNet. Start your own ZeroNet for complete experience. Learn More