Why Byblos

Compared To Existing Tools

Existing Tools Solve Pieces Of The Problem

People already have ways to work around the browser's limitations. Links are shared in WhatsApp groups. Discussions happen in Discord. Notes are written in Notion or Obsidian. Interesting pages are saved to bookmarks. Comparisons end up in spreadsheets. Tasks and reminders live somewhere else.

None of these tools are bad. In fact, most of them are excellent at the jobs they were designed to perform.

The reason people adopt them is that the browser provides very little support for evaluating information over time. Once a task involves comparing many webpages, discussing alternatives, recording observations, and gradually arriving at a decision, people naturally begin introducing additional tools to compensate. The result is a workflow that spans multiple applications, each responsible for a different part of the same decision.

The Airbnb Example

Imagine a family planning a vacation. Over the course of a week they collect twenty Airbnb listings. Some are too expensive. Some are in the wrong neighborhood. Some have incredible views. One looks perfect until somebody notices a major road running directly behind the property. Another initially seems expensive but later becomes attractive after comparing it to similar properties nearby.

The links are shared in a WhatsApp group. Opinions are exchanged throughout the week. Somebody bookmarks a few favorites. Somebody else creates a spreadsheet to compare prices and amenities. A handful of properties emerge as serious candidates. A week later the family returns to continue the search.

The Airbnb listings are still there. The WhatsApp messages are still there. The spreadsheet is still there. What is difficult to recover is the relationship between them. Which property was the one with the amazing view? Which property did your spouse strongly prefer? Which one looked promising until somebody discovered the road noise? Which properties were rejected? Why were they rejected? Which ones are still under consideration? The information has not been lost. It has been distributed across multiple systems.

A more organized group can reduce some of this pain in exchange for more. They might use Discord instead of WhatsApp. They might create a channel for each property. Reactions can serve as ratings. Pinned messages can summarize conclusions. Bots can track reminders and metadata.

This approach solves some problems and creates others. The Airbnb listing still exists inside the browser. The discussion exists inside Discord. Ratings exist as reactions. Reminders exist wherever the bot stores them. Search operates differently in each system. The relationship between those systems must be maintained explicitly. Nobody wants to stitch together a browser, a spreadsheet, notes, reminders, and Discord channels just to make a damn decision.

Page-Centric Tools

The limitations of browser-centric workflows have not gone unnoticed. Over the years, many tools have emerged to solve specific pain points of the workflow.

Pocket helps users save pages for later. Hypothesis and Diigo allow users to annotate webpages. Zotero helps researchers collect and organize sources. Bookmark managers such as Raindrop provide richer organization than traditional browser bookmarks. Other tools focus on tab management, note-taking, search, collaboration, or research workflows. These tools are often excellent at the problems they were designed to solve.

Their existence demonstrates that people repeatedly find themselves needing more than a webpage. They need notes attached to pages. They need discussions attached to pages. They need ratings, reminders, relationships, organization, search, and collaboration attached to pages.

Most page-centric tools address one of these needs at a time. Annotation tools focus on annotation. Bookmark managers focus on organization. Research tools focus on source collection. Collaboration tools focus on discussion.

The challenge is that these capabilities become most valuable when they interact with one another. A bookmark becomes more useful when it carries discussion, ratings, and reactions so you know why the page was saved. A rating becomes more useful when it influences which page you discuss next. Navigation becomes more informative when it reveals patterns in what people are thinking. The value of bookmarks, annotations, discussions, and navigation graphs are amplified by the presence of each other.

Byblos

Byblos brings these capabilities into the same workspace and treats them as parts of the same process rather than independent tools. As a result, Byblos can answer questions and make observations that many tools miss in isolation. Consider our Airbnb example from earlier:

• Which listings with keyword "pool" did Alice and Bob both like, sorted by cost?
• This listing passes my filters, why did we reject it? Oh, Alice left a chat.
• Why are the three of us looking for listings in different parts of the city?
• Alice is writing distances on pages, so she prefers listings in close proximity to events.
• I just joined the search, what is everybody thinking?
• How have people's preferences changed over time?

None of these questions belong to a single category such as bookmarking, annotation, discussion, navigation, or collaboration. They emerge from the interaction between all of them. Byblos exists because much of the value of page-centric work lives in those interactions.

Since the entire search and all relevant context is in the same place, sharing it with others is trivial. For example, Carol can join the Airbnb search with a single document instead of a folder of contextless tabs, a bookmark manager, a Notion document, a spreadsheet, and a WhatsApp thread.