Window/tab merge

Why do you have multiple tabs open? The answer is that you're cross referencing in your head. Why not let the computer do it?


What if you could cross reference websites automatically?

Essentially form a join between different sites of data.


投票 (可選) (別通知) (可選)


Would this automatically make view of tables of aggregated data from all the open tabs? I think we do need this, and not just from open tabs, but from entire browsing histories.

我懷疑這一切都可能是惰性生成的。每個選項卡都提供了一個用於數據訪問和檢索的 API,該 API 映射到站點以及屏幕上的內容。瀏覽器還將爲歷史和下載以及歷史頁面內容提供一些名稱空間。

I suspect it could all be generated lazily. Each tab provides an API for data access and retrieval which maps to the site and what's on screen. The browser would also provide some namespaces for history and downloads and historical page content.

    : Mindey
    :  -- 
    :  -- 


當然,這需要有人學會將任意頁面信息解析爲信息量合理的表格形式。我想出了一種格式,即“元格式”來表示這種任意數據源本體規範化,它可以很好地表示機器學習任務的訓練集,以創建自動執行這種表格化的模型。然而,如果沒有這樣的 ML 模型,我們將不得不讓人類編寫解析器並對每個頁面進行本體規範化,這非常耗時。

Of course, this needs someone to learn to parse arbitrary page information into reasonably informative tabular form. I had come up with a format, namely, "metaformat" to represent such arbitrary data source ontological normalizations, that could work well to represent a training set for a machine learning task to create such model that does such tabularization automatically. However, without such ML model, we would have to have humans write parsers and ontologically normalize each page, which is prohibitively time-consuming.

Linux 具有用於通用進程間通信的 DBus,而 gconf-editor 像編輯器這樣的註冊表使用 Dbus 來詢問可用的配置鍵。


Linux has DBus for generic interprocess communication and gconf-editor a registry like editor uses Dbus to interrogate available configuration keys.

Page normalisation will be a problem for on screen information. Would need to clean up data before piping to cross referencing systems.

    : Mindey
    :  -- 
    :  --