1. Configure PyWebScrapBook and set server.browse to false. We'll consider use false as default in the future.
2. The primary feature for WebScrapBook is capturing, not management, and thus putting "Open ScrapBook" as the first command doesn't make sense. Emphasizing which command is likely controversial for the same reason. We need a thorough evaluation/discussion before doing this. FYI: you can configure a hotkey using the browser extension hotkey manager for a faster way to open the sidebar.

This is current not supported and is an open issue: https://github.com/danny0838/webscrapbook/issues/8

If you have some good idea welcome to provide a further feedback in the thread.

煩請提供頁面網址及螢幕截圖，說明下是哪個頁面哪張圖有問題，以便除錯，謝謝

Update 2020-10-13:
https://www.yuque.com/yuque/help/dive-into-yuque-editor 此網頁內容是動態加載的，需要先把螢幕向下捲，讓整個頁面全部載入，之後再擷取並未發現圖片擷取有異常。

在這裡更新評論時我們不會收到任何通知，若要進一步討論建議回報到版本庫或寄電郵，否則我們可能很久才會發現及回覆。

WebScrapBook is not meant to be compatible with ScrapBook, whose data scheme is relatively old and is rather limited (e.g. can't support .htz, .maff, and single html). We may implement a tool to (unidirectionally) convert WebScrapBook data into ScrapBook-compatible format like ScrapBook X Converter, but likely being a command line tool using PyWebScrapBook for better performance and cross-platform compatibility.

For now, metadata like URL source of the captured page is saved in the index.html as a root HTML tag attribute, and can be easily read using PyWSB backend. You can manually (or write a script to) create an index.dat from it if you need to back-port a page captured by WSB to SB.

1) This is a current open issue (https://github.com/danny0838/webscrapbook/issues/37). It's technically possible but not quite easy to implement.
2) I don't know what the folder selection is for. Generally speaking, implementing a virtual folder tree in the context menu is something technically possible but not easy to do.
3) Automatic updating the sidebar after a capture may cause loss of current UI status such as selection, scrolling position, etc, and may largely increase network traffic if the backend server is hosted remotely. We need a good solution for such related issues before implementing this.