loom, the information browser.

loom is not a web browser - for some people, it's something better. it is a research tool - a document viewer - and an environment for research, learning, and working in a distraction-free environment.

loom is a browser that focuses on information, rather than content. to achieve this goal, loom prefers informational and direct hypertexts and documents, rather than feature-filled web applications and “experiences”.

to achieve this goal, we've made a few opinionated choices on the types of content that loom will fetch, display and make usable as information:

• ad blocking, served by community element and server filter lists, is used by default to extinguish attention-robbing third-party content from the information you seek.
• information transformation rules, contributed by the community and vetted by project stewards and other community members, can pull and present information from pages directly, for use elsewhere.
  • these same rules also influence the reader mode features.
• reader mode as a default. pages are harmonized to fit a view best served for displaying and reproducing information.
  • Pages are given the @media print media type by default. When in dark mode, content will be presented as if in printed page mode, but images will not be inverted.
• automatic local archiving. loom will, as you browse information, cache it locally, alongside the media and assets on the page.
  • as an option, loom can fetch/create remote archives through services including Internet Archive's Wayback Machine, and archive.is. This is useful for archiving sites “as-original”, rather than a transformed or limited version as loom presents.
  • Remote archiving can be set to be automatic for certain domains, but we strongly recommend to take care when using this mode, not to archive information or pages you might consider personal, secret or forbidden to disclose publicly.
• loom has a strong preference for simplified HTML and CSS, smaller file sizes, and less complex web pages.
  • The HTML/R reduced subset of HTML is used by default.
  • Scripts larger than a handful of kilobytes are forbidden.
  • Media is handled externally. <video>, <audio>, and similar tags by themselves, will work as expected, but will download and play the video in your chosen video player - not embedded.
  • Styling and CSS is supported and encouraged, but may be stripped or transformed for readability when in “reader mode”, as it is in other browsers.

You can configure most or all of these functions to your taste, but some high-complexity, high-power or potentially dangerous things considered necessary for web standards compliance, may be considered anti-features to the loom project, and might not be implemented. we consider this a good thing.

the loom philosophy is that of the earlier Internet and the World Wide Web - that of POST and GET, of pages being rendered and displayed once. today's internet is fast enough that a lot of concessions made in the 90s and 2000s for dial-in users don't make sense today, and as a result, a lot of information design and practical web publishing has lost the information-capital, information-first nature they sought to serve in the first place.

information now takes a backseat to interactivity, to pulldowns and hamburger menus, to JSON and AJAX and JavaScript and application-scale web “experience” development writ large, and served by “no-code” frameworks which can't know whether you're posting a small blog, or reservations for a local board game night, or an entire operating system-in-a-browser focused exclusively on the provision and serving of commingled but completely unrelated community experiences under a single 'brand'.

all that misses the point, we feel, of a dedicated application for retrieving information. gopher and other early Internet protocols tried to fit this niche, and Gemini today is doing a neat thing, but realistically, far too much information exists in the form of the world wide web as it is today - and you can't be expected to teach people gemini and have everyone make both a “web version” and a “gemini version” of everything.

so, loom desires to take this existing information and present it - without distractions, without games, without “subscribe now”'s and “free trial”'s.

loom can fetch pages of information from many protocols, not just HTTP/HTTPS. we anticipate supporting any useful protocol where a 'file' or 'record' can be retrieved and presented, including file:, FTP, SFTP, DNS (as records), veilid, tor, gemini, ipfs, http/r as well as traditional http1/2/3.

when you visit a website, by default, the site - and anything directly included to it - is downloaded and kept, not just “cached”. it is then presented from this local version. you can copy it out and modify it, too - loom will tend to track and update changes to the version it had fetched, with all changes tracked in a version control scheme (like Git).

loom will surface bookmarks to informational, reference and educational websites, which follow the loom philosophy - even when not presented through loom itself. these bookmarks are vetted by our community, stewardship and release committee, and presented when installed, and - at the user's option - with occasional updates. sites may be removed or added to this directory regularly, and websites can qualify and disqualify themselves a week at a time.

user submissions for the ad blocking filters, and information transformation, and other user-submitted stuff, is intended to be licensed under a permissive license. loom itself is to be licensed under an open source, free-software license, and provided at no cost to the user.

• Q. What does 'loom' mean? Where does it come from? Does it have anything to do with [other things called 'loom' or which rhyme with it]?
• A. a loom weaves fabric from lines of cotton. loom weaves pages of information from lines of text. there's probably “warp”, “weft” and “shuttle” somewhere in there.
• A. “loom” may or may not stand for Library Of Our Minds?