+ Lots! Packrat lets you export your archived pages as WACZ files
+ which are supported by various web archiving tools such as{" "}
+
+ Browsertrix
+ {" "}
+ and can be viewed or embedded in your own content using{" "}
+
+ ReplayWeb.page
+
+ .
+
+
+ WACZs are also just ZIP files. If you unzip them they'll reveal
+ their component{" "}
+
+ WARC files
+ {" "}
+ (another web archiving format) that is compatible with other
+ tools.
+
+
+
+
+
+ What analytics information do you collect? How do you use this
+ data?
+
+
+
+ If analytics are enabled we collect the following anonymized
+ data:
+
+
+
+
+
Tracking Identifier
+
Description
+
+
+
+
+
ViewPage
+
+ Triggers when a user clicks to view an archived page.
+
+
+
+
TorrentCreated
+
+ Triggers when a torrent is created for sharing archived
+ content. Tracks how many pages are included.
+
+
+
+
PageArchived
+
Triggers when a page is successfully archived.
+
+
+
PageSize
+
+ Logs the size of the archived page if available, rounded
+ to the nearest kilobyte.
+
+
+
+
CookiesEnabled
+
Tracks if cookie archiving is enabled.
+
+
+
LocalstorageEnabled
+
Tracks if local storage archiving is enabled.
+
+
+
ScreenshotsEnabled
+
+ Tracks if thumbnail screenshot archiving is enabled.
+
+
+
+
AnalyticsEnabled
+
Tracks if analytics tracking is enabled.
+
+
+
SkippedDomains
+
+ Logs the number of domains added to the skip list when
+ changed.
+
+
+
+
TotalSize
+
+ Tracks the total size of all archived content, rounded to
+ megabytes.
+
+
+
+
Started
+
+ Triggered when the archiving process begins for a page.
+
+
+
+
Stopped
+
+ Triggered when the archiving process is stopped, either
+ manually or for another reason.
+
+
+
+
+
+ This information helps us understand if you're finding the
+ features we've developed valuable! We self-host our analytics
+ and we will not sell or reveal this data to third parties.
+
+ We believe that browser history and bookmarks are some of the
+ most un-iterated-upon features shipped in browsers today. In
+ recent years other companies have stepped in to help fill this
+ innovation void:
+
+
+
+
+ Obsidian has released{" "}
+
+ Web Clipper
+ {" "}
+ which enables users to extract key text and information from
+ pages they visit as Markdown to save in their own notes.
+
+
+
+
+ Webrecorder maintains{" "}
+
+ ArchiveWeb.page
+ {" "}
+ which allows users to create focused, curated web archives
+ of content they care about. ArchiveWeb.page is great, but
+ like Web Clipper it's ui is focused on the user making
+ active curatorial decisions about what content they deem
+ valuable.
+
+
+
+
+ Other tools like{" "}
+ ArchiveBox require
+ users to host a dedicated server to handle extraction and
+ preservation of content — a much more cohesive record of
+ what was browsed, but at the added cost that comes with
+ running infrastructure.
+
+
+
+
+
+ We think that there is opportunity for a middle ground between
+ these tools. Building upon Webrecorder's work enabling fully
+ browser-based capture and replay of web archives, Packrat seeks
+ to offer the same ease of use afforded by ArchiveWeb.page but
+ with a new user interface focused less on curation.{" "}
+
+ Chrome's Side Panel API
+ {" "}
+ (launched in 2022) allows us to showcase the user's archived
+ pages directly beside the current page with the same Material 3
+ design language adopted by Chrome, enabling a seamless
+ transition between live web content and navigating archived
+ pages.
+
+
Challenges
+
+ Extensions aren't without their limitations, the biggest one we
+ face is our use of the Chrome Developer Tools API which, for
+ well-intentioned security reasons, displays a persistent banner
+ above the current page telling the user that an extension is
+ "debugging the page". While users can hide this by launching
+ their browser with the{" "}
+ --silent-debugger-extension-api command line flag
+ if they are sufficiently motivated to do so, ultimately we are
+ restricted by the browser environment.
+
+
+ Webtorrent is close to a perfect solution for hosting
+ decentralized web archives, but it still isn't supported in most
+ major torrent clients as of publishing. Unlike most other
+ standard torrent clients, Webtorrent uses WebRTC to communicate
+ between peers.{" "}
+
+ While WebRTC support was added to libtorrent
+ {" "}
+ (a common C++ library used by many torrent clients) in 2020, as
+ of publishing it is hidden behind a compiler flag and no release
+ of libtorrent has shipped with it enabled by default. In
+ practice, most clients follow the defaults of libtorrent and
+ therefore don't support seeding to Webtorrent peers. While this
+ doesn't pose a direct problem for Packrat — our peers are always
+ other Packrat instances using Webtorrent — having this enabled
+ in libtorrent by default would enable other interesting
+ decentralized web archive hosting projects.
+
+
+
+ We hope that you find Packrat useful, either as a case study or
+ as a tool!
+