Quotes in the clouds

We describe the concept of the interface of the context input of quotes. It will not be about the clouds, which everyone thought. Clouds are like the “Heaven” benchmark, the 4th version of which was recently released (fig. On the left) or the Sky Castle , and the quote pad is an interface element such as flying islands. On the surface of these islands, life is warming, and they themselves can fly. In our case - move the mouse. They are supported by stones, and we have contextual buttons.

How is a flying island made? Select the context with the mouse. Selection creates the basis of the surface - a quote. Under it appears a button, barely noticeable, translucent, which we used to see in the Word, starting with "2007". It may disappear soon, and correctly, if the selection of the text was created for other purposes. But if you hover the mouse - it creates the support of our future quote book, floating in the clouds.

Why do I need a quote? It turns out goals can be many. Quotes are needed if we want to respond to the statement in the article. If you want to remember a quote. If you want to report an error in the text or discuss the quote in the letter. Even a comment on an article is often more convenient to write next to the context, like everything else. This is how a flying castle is born - light and compact, instantly movable and reappearing in one movement of the mouse. If it was drawn in the film, then why not draw in HTML?
')
The list of quotes and corrections created in the article is displayed as a list of links above the citation block. The letter of the reference displays the type of quotation (quote, logical error, grammatical, punctuation, typo, or stylistic). They correspond to the characteristic colors for each type of corrections. The colors are needed by the proofreader (the person creating the quotes) in order to quickly navigate the set of quotes:

C - C itatata - prevails in discussions;
w - o logical logic / actual. Rarely encountered;
G - G rammatical - the most frequent;
- nonctual - second in frequency;
O - O signet - 2-3rd in frequency;
C - C tilistic - rarely observed.

For the most common grammatical and punctuation errors, the most contrasting, distinguishable colors with associations by marking in Word are selected. Places of errors of logic and stylistics in quotations stand out longer, therefore for them - the colors are quieter.

The method of entering errors, their places and replacement texts is made as “lazy” as possible, using a minimum of operational elements. Special error fields are not created. Users are not forced to rewrite or copy text or even read 2 copies - incorrect and correct. Instead, they simply click on the error type button (or the quotation button) first. There are six such buttons. A new quote icon is created above the quote and placed in a row of previously created characters in the order in which they appear in the text.

There may be one or more errors in the quote. Therefore, the letters of the types of errors are combined together and look like a syllable or a word: how many letters, how many errors are noted in one quotation. And if the quote has errors, then the letter "C" is not required for it.

How does the quoting process

1) Allocate a quote for commenting or with one or more errors. The algorithm will automatically capture partially captured words in the quote and add a line before and after, if the quote is short (less than 1 line);

2) Press the context button - type of error (1 of 5) or “comment” ("<_>"). The selected text will immediately form a quotation - a text without formatting in a block of approximately square outlines. Above the quote appears a list of previously created quotes for this article. The symbol for a new quotation takes place among others in the order of the quotation in the article. In the same list - "View quotes to other articles," if there are any in the repository.

3) Under the quote appears an empty input field for corrections or comments. For a comment, simply enter your text (the input field will grow). For correction - in the quotation highlight erroneous symbols. It is permissible to capture neighboring ones so that the selection in the letter is noticeable.

4) After selection, a copy of the selection appears in the input field. These are the symbols that are then replaced with the correct one. If it is necessary to say that these symbols are superfluous, they are deleted. If you need to correct or add - insert their characters. Everything, you can repeat steps 2-3-4 again to describe another mistake or comment.

5) All characters in the input field are immediately remembered as an edit. No need to press buttons, enter Enter or Ctrl-Entrer. To complete the editing operation, it is enough to exit the context entry of the comment by clicking, including randomly, on a different window location. It is easily continued by choosing from the list above the quote. The current quotation is remembered, return to it is performed by the return button: "<_".

6) When you return to the edit or when choosing the previous edit, the whole context of the quotation is restored: quote, comment and all errors (replacements) noted in the quote.

7) To correct a comment to a quote, simply select the quote again and rewrite the comment. To correct the contents of the error, click on the error area and correct the replacement text. To remove a quote or mark an error in it, a gesture is used - throwing a quote from the list or area of ​​error from the quote. Neither quotation boundaries, nor error boundaries can be corrected; you can only delete them and recreate them.

So, naturally noting the mistakes, the proof-reader does not think about sending a letter until the intuition tells you that it is time. No problem. Quotes are stored in the repository, so they are accessible from the browser on the computer all the time, from any window, with possible shutdowns of the browser and computer. They are constantly visible in the edit context box. When switching to other articles, there is a button to display the entire list of saved quotes.

Mark quotes in the article

If necessary, the quotes look like sections of text highlighted in the (uncorrected) text of the article. All of them have at least 1 character, so the link is placed on the site, clicking on which returns to the error editing. For the convenience of clicking and future reading, 2-3 characters around the error are automatically highlighted if the selection takes 1-3 characters. Visually, the selection looks like an underline and a paler selection of a text area with a background color corresponding to the type of error.

To transfer quotes to another browser or computer, click the Export / Import button. Therefore, if the dialogue does not require an immediate start and a quick send, it is possible to send written from another computer.

If the author of the article accepted corrections and corrected exactly, the quotations are no longer found in the previous form, but are searched for in a new one - they are marked as accepted and removed from the article. If they are no longer searched for at all, they are marked as “lost”, and the place of their marks is marked by highlighting the background before consideration.

For statistics and monitoring, the script is charged with counting the number of types of errors and reporting these statistics briefly at the beginning of the letter.

Way to send quotes

Old fashioned way. Encoding will not be visible to the addressee. He will receive a letter in a polite form representing the location of errors and possible corrections. In the final letter, the colors of the errors will only illustrate their place in the quotations and help to visually find out.

Modern The author receives a list of edits in JSON, which are displayed in his article also as text selections. In the contextual quotation, a list appears not only of its quotations (if any), but also of those who arrived. To accept them, the "accept" button is pressed (the whole quote or individual errors). A floating frame is opened with the article editing form, and the specified quotation or error is searched for in the text, and is rewritten to the replacement text.

If the author partially agrees with the replacement, he changes the text of the correction. The commentator is answered with a change. If he does not agree with the replacement, throws it out with a mouse gesture from the quote. In any case, the commentator comes back.

Comments to the quotation (including those with errors) come in the form of texts after the quotation. The author has the opportunity to edit not only the quotation of his text, but also to correct the comment that came, including an indication of errors. Write your comment - it means to continue the discussion. It creates, in fact, a comment thread on the quote. Comments of quotations are “accepted” by the same button, which in this case means “read”. There is no action on the text, only a message is sent to the commentator. If several “emotional” buttons are envisaged, a mechanism is created for evaluating the quotation or commentary on it.

(About this method of commenting somewhere long ago they wrote and even made an implementation, offering to highlight the comments in yellow. They also joked in the answers that in time the whole article will turn yellow.)

Discussion repositories

To the futuristic interface of discussions we will finish offtopic: infrastructure. A commenting mechanism is created, tied to quotes, a mechanism for correcting and discussing errors and quotations, and it is not necessarily tied to the author of the article or a specific site. The addressee of the commentator does not necessarily indicate the author, but, for example, the circle of friends or colleagues with whom an arbitrary page is discussed. All text of interest and discussion flow is stored in the memory of computers of interested parties. The button to respond to the initial submission of an article is an assessment that speaks publicly whether the addressee will remember this page and participate in the discussion on it.

As a result, the system is obtained from a mixture of Digg, Reddit, Disqus, and regular forums and social networks. Yes, even with a mixture of P2P, which is significant. On the side of the discussion, sources of information are not involved in principle. They are required to feed the source once. If the server-centered media users depend on the policy of the sources, then here they themselves manage the sources. Sites are trying to highlight the right audience, disperse the wrong. Here this task is assigned to the groups themselves. Each person is valuable with his own knowledge and the ability to process others.

On the side of the source of articles, there remains a passive reception and registration of discussions, if the source is interested. If discussions are not registered, all their preservation rests on the interested participants. Participants gain control over what they read and their messages. If the reader is interested, he remembers the article in a personal archive. Personal archives work offline, with the loss of the main source distributed by p2p.