almost everything from the outside world now comes to us through our web browsers. we see it, we forget it, but sometimes it leaves an impression. what would it be like if we could recall it at any time, just as we saw it?
even though the web is constantly changing, even though web pages and entire hosting sites disappear, you still get to keep what you saw with irchiver. seeing is keeping
irchiver aims to be your backup memory. just like how the internet has changed the way we think about knowledge, does having our memory on-demand change how we remember? irchiver is a long-term research project to study this question
the irchiver software is available for free, and it currently runs only on Windows 8.1 or higher (but it has only been tested on Windows 10) with five popular web browsers: Chrome, Firefox, Edge, Brave, and Opera
MacOS is not supported, but may be possible with a browser extension that mirrors to a Windows computer; no development is planned for MacOS due to limited developer time and expertise, and the inherent challenges of finding techniques for cross-process window access that continue working long-term
irchiver does not use any network connectivity for your privacy. why would you want your browsing data stored on the cloud? irchiver saves your screenshots as webp files (and png files temporarily), and the text content in a plain text file, in a folder on your hard drive. you can view it using any file explorer, and do whatever you want with it
| product | when captured | stored | format | reproduction | platform |
|---|---|---|---|---|---|
| irchiver | continuous, automatic | local | image, text | what you saw | Windows, any browser |
| archive.org | occasional, automatic | cloud | webpage | only public view | web |
| archive.ph | manually initiated | cloud | webpage | only public view | web |
| ArchiveBox | manually initiated | local | webpage | only public view | any desktop |
| WebRecorder | manually initiated | local | webpage | what you scroll | Chrome |
| taking a screenshot | manually initiated | local | image | what you saw | any desktop |
| browser extensions | manually initiated | local | varies | varies | varies |
| browser "Save as PDF" | manually initiated | local | nearly what you saw | most browsers |
irchiver is the only way to reproduce what you actually saw with continuous automatic capture. other products either require manually initiating a capture each time, or can only capture stateless publicly-accessible pages. irchiver saves captures as images and plain text, as historically those formats have outlasted software code and cloud services
this is the current roadmap for irchiver as of 2021-12-02, but to hear about releases, please subscribe to be notified about updates
version 0.2: 2x more efficient (cpu/memory performance-wise), while consuming 2x less space on disk; enable archiving options that are currently inactive in the context menu
version 0.3: better installer and more install options; notification tray icon indicator during capture; an additional 2x more efficient (cpu/memory performance-wise), and further reduction in space usage
version 0.4: API with support for python extensions/plugins, for example: automatic scrapbooks, highlight page changes, etc.; an additional 2x more efficient (cpu/memory performance-wise)
version 0.5: better search interface and a by-date browsing interface; partial-match text search; better text recognition in images (e.g. infographics, webcomics)
irchiver is licensed to you under the Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) license, with the additional condition that it can only be used with the enthusiastic consent of the person using a computer it is installed on
while the indexer and search interface are open source, irchiver's capture software is not open source in the traditional meaning of the term, but rather it's source-verifiable, where you can watch the main executable file being created and the SHA-256 hash generated for the executable so that you know the distributed executable is from the source code seen in the video
the capture software is not open source to make it harder to deploy the software without the user's knowledge, which can be done by modifying the source code to hide the notification tray icon and thus rendering the automatic background capture completely invisible
feel free to email me to request the capture software source code with an explanation what it will be used for and how it will be distributed, or for requests for other licenses