Wiki: The FYWD Browser
Web design isn't the solution. Web design is the problem.
Site-offered CSS would be ignored (or at best strongly deprecated). Straight-bog HgsML 1, 2, / HTML5 sites would render best. User could opt for specific client-side stylesheets, e.g., night mode, large/small font, etc.
At your option: "Fine Young Western Dinosaurs", or "Fuck Your Web Design". One of these is more fundamental to the concept.
- Present information and just the relevant information.
- Minimise any and all distractions.
- Provide a useful platform for acquiring, organising, annotating, and utilising large volumes of information
This is the bastard chimeral love child of FBReader, Pocketbook, Calibre, Pandoc, Zotero, and Readability / Pocket.
While Ello deviates slightly from my preferred layout standards, on whole it's far closer than many websites to the ideal, and the underlying HTML isn't bad.
The "Edward Morbius's Motherfucking Website" CodePen (see below) is close in general concept to a non-distracting, and flexible, display.
- Sane-by-default HTML entity attributes. W3C defaults are not sane.
- Flexible formatting across a wide range of display properties, formats, and devices.
- A set of predefined page layouts which are recognised and supported by default. Other formats may be presented, but less optimally.
- Distraction minimisation as a key design concept.
- Support and preference for enhanced metadata formats, e.g., hNews.
- Local / own systems management of resources, not cloud-based, except by user preference.
- Bibliographic data support. Essentially BibTeX DB creation, import, and access.
- Tagging and classification. Both standardised and custom classifications. For research, something close to Library of Congress Catalog System classifications, if not exactly that.
- Citations, both ways. What does this cite / where is this cited?
- Persistent local content cache.
- Searchable local corpus.
- Author/publisher identification and reputation management. Identify who is and is not a reliable or trusted source.
The default templates are a set of page layouts which can be 1) used by authors and for which 2) a set of default (and extensible) client-side stylings will be provided. This is very much open to suggestion. Think in terms of the base page styles of a good CMS. See for example Drupal's site building page, and keep in mind I'm not particularly versed in Drupal.
Article. Straight-up text. Supports metadata (title, author, publication date, summary/abstract), inline images, structural markup, notes (foot / end / side / hover). May include a discussion thread.
Discussion thread. Why this doesn't exist as a native, supported, document component within HTML is beyond me. With the first interactive Web pages, numerous sites essentially reimplemented the concept of a threaded (or flat) Usenet discussion. See below for features, but generally, various threading models, filtering, expand/collapse, summary/full, etc. Each subthread should be a its own individually addressible HTTP target.
Index page. Generally, a page which references other content, is dynamic, and presents two or more primary content items within it, as well as other fixed elements and/or callouts. A typical "front page".
Gallery. Similar to an index, but with a larger focus on image or graphical presentation.
Directory. Similar to an index page, but in which components are more static, generally pointing to other parts of a general collection.
Image. A page on which the principle focus is to present a single image, or a set of related views, generally one-at-a-time, though multiple on a page might also be used. Little text, that largely contextualising the image.
Table. A tabular, spreadsheet, or workflow type layout. Should have options for interactivity, including show/hide rows, column sort, and column filter. Useful for, say, a ticketing or events system.
Data. Presentation of data in both tabular and graphical form. Should offer some degree of interactive support. I'm thinking along the lines of SVG, D3.js, Plotly, etc.
Multimedia. Generally, any substantive multimedia content -- more than ~10s of audio or video, should be directed to a specific multimedia application with queing and dedicated controls. No media shall be allowed to autoplay.
Application. (Deprecated.) A page in which primary functionality is to provide some form of interactive application.
Unstructured. (Deprecated.) A page with no inherent internal structure.
There are numerous websites (including Ello) whose essential function is to replicate some subset of Usenet or BBS style discussions, though sometimes with additional features. This functionality and the entity primitives isn't built into the stock HTML definitions, but almost certainly should be. Doing so would allow a client rather than a website to provide for core functionality including:
Display mode. Flat, threaded, nested, FIFO, LIFO, expanded, collapsed.
Rating and reputation. Marking content for quality, relevance, interest, factual basis, civility, etc., and recording on an author basis the long-term reputation of a given source.
Filtering. Rating, reputation, and content mean little without the ability to suppress content of little or no interest. This would include the ability to block specific authors and/or topics.
Sortition. One of the best articles I've read in the past five years has been Michael Schulson's "How to Choose" in Aeon. There's been a tremendous amount of focus in filtering systems to present precisely assessed items in rank order. Rather than this, bucketing content by quality/relevance measures, and presenting some subset of low-ranked material along with highly relevant material, would get past many problems of "how do you start introducing previously unknown content", without overwhelming all users with crap, or without arbitrarily suppressing all but a small quantity of low-ranked or unknown content. A random presentation breaks this deadlock, and allows for a reshuffling (by resetting the random seed or basis) to surface a new set of content.
Peer recommendations. An issue I've had with blocking is that occasionally an otherwise-blocked source turns up something of interest -- either for quality or extraordinary lack. In which case, promotion by a known peer might be of value.
Pseudonymity / anonymity. Though authorship/authority is considered baked in, an "author" is merely an asserted identity, which may be persistent (pseudonymity) or one-time (anonymity). The principle difference is that a pseudonymous author may achieve a reputation whilst an anonymous author has none. An anonymous author might, though, be vouched for by another (pseudonymous, anonymous, or "real name") entity. See the "Yale Wall" concept from Lawrence Lessig's Code and Other Laws of Cyberspace.
A possible content-ranking scale
The system I've been employing on Pocket is to define a set of tags I'm using to describe article quality: a-0, a-1, a-2, a-3, a-4, a-5, ... Currently 'a-5' is the highest level, a-0 is the lowest.
The general allocation is:
a-5: Content is definitional in its field. This is the best current state of knowledge. Largely reserved for articles or scientific papers, may also apply to specific legal documents (e.g., the US Constitution, the UN Declaration of Universal Human Rights). Darwin's The Origin of Species. A general interest article (such as Schulson's article on choosing, above) may end up here if it particularly illuminates or raises an interesting point.
a-4: Very high-quality documents. Clearly or authoritatively discusses a topic. A well-written academic text will reach this level. Requires clarity, originality, substantial documentation (footnotes/endnotes, bibliography, and index). Some articles, the very rare blog.
a-3: Quality presentation of a concept or topic. A lay book, a particularly well-written news or magazine article, a very high-quality blog article, will generally reach this level. Many poorly-written technical articles whose primary goal is meeting publication and obfuscatory obligations reach only this level. Many of The Intercept's articles, or Pro Publica or the ICIJ's series on offshore banking are in this classification.
a-2: Informative but not exceptional. Most better-quality news coverage in a major national paper (The Guardian, The New York Times, The Economist, tend to be about here. The information is solid, there is some background provided, but the contextualisation is at best middling. Many stream-of-consciousness blogs hit this level -- blather without illumination or support.
a-1: Not obviously false reporting. The bulk of news coverage that serves largely to document that some specific thing occurred, but fails to contextualise or illuminate beyond the level of simply journaling incidents. Also a great many blog entries.
a-0: You're stupider for reading it. I don't classify much in this category, as I'm generally not interested in vast compilations of stupidity, but the occasional odd lot can be useful. Authors finding themselves here generally have their credibility seriously impeached.
There's a tremendous amount I don't classify, of course, and some (though not all) of that would fall under a negative level classification.
I've also thought of creating an upper-level classification, a-6, for materials which introduce a (currently) entirely novel concept. Strictly, works such as Newton's Principia, Darwin's Origin, or Claude Shannon's information theory papers should be here, though they now largely belong to history as opposed to current development. I may move them to this realm.
Note that this is an index of conceptual significance which may not apply to other types of works, say, literature.