ZeroNetX / ZeroNet

ZeroNetX - Decentralized websites using Bitcoin crypto and BitTorrent network
https://zeronet.dev
Other
212 stars 32 forks source link

Big files (10MB and more) are not synchronizing #153

Closed defder-su closed 1 year ago

defder-su commented 2 years ago

Step 1: Please describe your environment

Step 2: Describe the problem:

Steps to reproduce:

  1. Create site (don't use optional files).
  2. Add some big file (e.g. 15MB zip).
  3. Add some middle file (8MB zip).
  4. Sign and publish the site.
  5. Visit the site from some other ZeroNet node.

Observed Results:

Expected Results:

defder-su commented 2 years ago

Big optional files work (ZeroNet splits them automatically), but optional files are not durable.

defder-su commented 2 years ago

Reproduced it with http://127.0.0.1:43110/1Pj7TzBUdQRk9T9mzPQBjv6ju5oBiLCtry/ (file archive). Almost all files (more than thousand) are successfully downloaded, except 35 big files.

mx5kevin commented 2 years ago

This is a small part of the problem

-Not posible to download all large files simple way when a user download a single user zite.

-The incress site limit question: even if it is highlighted, larger and all optional files will not be downloaded. If not incressed, the page will not display correctly.

-Using one click the system is unable to download the files, does not finish downloading files. If B the user From A the (site owner not get the full content), C user from B will not possible to get the content. And it won't be C, D and E seeders conception. This system is unable to finish downloading a file. The download will be interrupted for all larger files.

The P2P protocol and the engineering solution conflict. Until this network becomes Seeding centric where the most important From A the (content owner) B the user get all content. And user C get from B without A if a goes offline the content. Until then, this network will not work.

mx5kevin commented 2 years ago

The problem is that the original developer did not see this network through the eyes of a website owner who using a PC, Laptop, and not a 24/7 running server. It looks like haven't been tested regularly in a live environment how it works with test pages. Anyone who sets up a website and tests from another network as user immediately see the serious problems. If the point B not get from A the content then how do you get it C from B? The other from the user's point of view is simple download. Most people can’t handle more complex options and don’t understand what it is. In single user zites download the full content default way (if someone doesn't need for the optional files than can turn it off if more competent, but the average user should not be disturbed by this). In multi user zites download the current content if the user click the same file. If someone wants to start a video, download a file don’t look at why it’s not available.

defder-su commented 1 year ago

PR https://github.com/ZeroNetX/ZeroNet/pull/204