![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Annotation. The paper discusses methods for creating social networks that will be simultaneously free from censorship and spam by the architecture. The proposed architecture is total and universal. At the user's choice, the system can look like a social network, a blog, a news feed, Wikipedia, or a large annotated Web directory suitable for searching the entire Internet. The subsequent texts discuss the influence of communication systems on human mind.
A social network, and not only social networks, is a combination of multiple different objects:
A. Database of texts with metadata and links,
B. Mutual relations of users including discourses , i.e., another database,
C. Hosting – synchronization – routing for A and B,
D. Frontend. User experience depends how the database A requests are set in the frontend. A social network or something else can be shown to the user according to the frontend settings.
Each of these objects is quite independent from the others. Each object sets its own direction of development: exploring relations of texts, finding user relations, hosting and viewing methods.
Each text can be described within “orthogonal” (independent) coordinates[2]:
1. The title, like in Wikipedia, is the “largest” coordinate. Wikipedia categories and reciprocal links allow to quantify the distances between texts.
2. The discourse of the text[1], which is just a point of view. The most important parameter the public is shy about on Internet. As explained below, the explicit statement of one’s point of view will allow to eliminate spam and censorship.
3. The author(s) of the text.
4. The dates of publication and editions like in Wikipedia or Facebook.
5. The language as shown in Wikipedia in the left column.
6. The level of complexity from school kids to professionals.
7. The format, such as encyclopedic article or original paper, forum, blog... Wikipedia has not only encyclopedic articles, but also forums for each article.
8. The significance and scope of the text, the types of external resources corresponding to the text, including third-party Websites or physical objects. For example, discussing a bakery chain or a specific bakery, which also applies to 1 and 7. The legal status of the text, for example, prohibited by law in certain jurisdictions. This coordinate is not completely independent, and is a set of coordinates rather than a single coordinate.
On Wikipedia we navigate the titles or the languages. It is easier to navigate, search and link by changing only one coordinate with all other coordinates preserved. Selecting the current coordinate for navigation sets the shape of the system. For example, if we navigate through the titles, the system is similar to Wikipedia or another Wiki system. Fixing the discourse sets the point of view of the wiki system. Navigating through the authors turns the system into a social network if the publication time is recent. The authors can be selected within a certain discourse or simply friended. For the news feed, it is also necessary to set recent dates, the format - original news publications, the significance - high and navigate the titles and the authors.
A new text has to be different from the existing ones at least one coordinate. Some coordinates may be missing, for example, a blog may not have topics, but there must be an author, and encyclopedia article will not have an author, but a lot of authors in the history.
Likes and dislikes might indicate just quality of the text or agreement with the point of view. Point of view likes indicate the same discourse and dislike indicate the other discourse. Such likes allow to calculate relations between the authors and distances between their discourses. Discourse is broader than friendship and can be calculated between the groups of users, and not just individually as a friend.
We require that the hosting can guarantee the absence of censorship and spam: any contribution is accepted, but we spam is hidden from us. For the absence of censorship, you need a distributed hosting with potentially unlimited number of owners. To avoid spam, we host only what we write ourselves and whom we read, our friends. We mark spam as spam, not a dislike, and trust our friends in assessments within the discourse. For routing and searching, we store ourselves and friends + a detailed map of our discourse (friends of our friends) + a detailed map of neighbors at least 1 step at each coordinate + an approximate direction where to look for what is far from us. A distributed single-ranked system cannot use user registration with logins and password and does not require it. Users create signatures to indicate the source of contributions. Others friend and judge contributions based on the signature.
Frontend is copied from Wikipedia, hiding languages other than languages we read, for example, only Spanish and English. In addition to languages, the left column navigates to other coordinates, including discourses. The navigation method specifies the type of resource we are reading. For example, a selection of recent posts by friends is a social network, navigation by article titles is Wikipedia, by discourses – viewing alternative points of view on one topic. Other forms are simple sets of rules created by users and loaded into the fronted.
Navigation using orthogonal coordinates places closely related texts side by side. For example, all the top articles about something from different points of view will be placed next to each over in the discourse navigation panel. The proximity of closely related texts helps to fight copy-paste. Of the similar publications, the most popular will have a higher priority. It also allows a quick response to other people's publications.
Search engines and Web directories promote monopolies, including the monopolies of Google as a search engine and Wikipedia as a large directory. In the mid-90s, there were directories of the entire Internet, then search engines (with a hidden Web directory inside), but of the large catalogs, only Wikipedia remained, all other Wiki projects are many orders of magnitude smaller. W++ is potentially a very large system, a monopoly. W++ can absorb any contribution that can be parsed into W++, not only all Wikis and all smaller clones of W++. Users prefer the largest system, the largest catalog of everything that creates full-fledged search functionality inside W++.
The manifest can be partially implemented, but full implementation will be the most viable, i.e., capable of absorbing partial implementations of W++ and the rest of the Internet.
[1] Discourses are dominant points of view shared by a large number of people.
[2] The coordinate system described below roughly corresponds to a multidimensional OLAP cube The coordinates are poorly ordered, but generally have a graph structure formed by references usually directed along one dimension. The application of OLAP to social networks is not new. Novelty of W++ is the access of users to navigation through OLAP indexes.
This is an author translation of https://vignatovic.dreamwidth.org/444.html from Russian to English.
The same text is also published on Reddit: https://www.reddit.com/user/vignatovic/comments/umqsam/w_manifesto_a_single_engine_and_a_single_database/
Bitcoin address to support the author bc1qrccvvnjmm33y8u6axsw007739vdzsnzw88znx8
A social network, and not only social networks, is a combination of multiple different objects:
A. Database of texts with metadata and links,
B. Mutual relations of users including discourses , i.e., another database,
C. Hosting – synchronization – routing for A and B,
D. Frontend. User experience depends how the database A requests are set in the frontend. A social network or something else can be shown to the user according to the frontend settings.
Each of these objects is quite independent from the others. Each object sets its own direction of development: exploring relations of texts, finding user relations, hosting and viewing methods.
Each text can be described within “orthogonal” (independent) coordinates[2]:
1. The title, like in Wikipedia, is the “largest” coordinate. Wikipedia categories and reciprocal links allow to quantify the distances between texts.
2. The discourse of the text[1], which is just a point of view. The most important parameter the public is shy about on Internet. As explained below, the explicit statement of one’s point of view will allow to eliminate spam and censorship.
3. The author(s) of the text.
4. The dates of publication and editions like in Wikipedia or Facebook.
5. The language as shown in Wikipedia in the left column.
6. The level of complexity from school kids to professionals.
7. The format, such as encyclopedic article or original paper, forum, blog... Wikipedia has not only encyclopedic articles, but also forums for each article.
8. The significance and scope of the text, the types of external resources corresponding to the text, including third-party Websites or physical objects. For example, discussing a bakery chain or a specific bakery, which also applies to 1 and 7. The legal status of the text, for example, prohibited by law in certain jurisdictions. This coordinate is not completely independent, and is a set of coordinates rather than a single coordinate.
On Wikipedia we navigate the titles or the languages. It is easier to navigate, search and link by changing only one coordinate with all other coordinates preserved. Selecting the current coordinate for navigation sets the shape of the system. For example, if we navigate through the titles, the system is similar to Wikipedia or another Wiki system. Fixing the discourse sets the point of view of the wiki system. Navigating through the authors turns the system into a social network if the publication time is recent. The authors can be selected within a certain discourse or simply friended. For the news feed, it is also necessary to set recent dates, the format - original news publications, the significance - high and navigate the titles and the authors.
A new text has to be different from the existing ones at least one coordinate. Some coordinates may be missing, for example, a blog may not have topics, but there must be an author, and encyclopedia article will not have an author, but a lot of authors in the history.
Likes and dislikes might indicate just quality of the text or agreement with the point of view. Point of view likes indicate the same discourse and dislike indicate the other discourse. Such likes allow to calculate relations between the authors and distances between their discourses. Discourse is broader than friendship and can be calculated between the groups of users, and not just individually as a friend.
We require that the hosting can guarantee the absence of censorship and spam: any contribution is accepted, but we spam is hidden from us. For the absence of censorship, you need a distributed hosting with potentially unlimited number of owners. To avoid spam, we host only what we write ourselves and whom we read, our friends. We mark spam as spam, not a dislike, and trust our friends in assessments within the discourse. For routing and searching, we store ourselves and friends + a detailed map of our discourse (friends of our friends) + a detailed map of neighbors at least 1 step at each coordinate + an approximate direction where to look for what is far from us. A distributed single-ranked system cannot use user registration with logins and password and does not require it. Users create signatures to indicate the source of contributions. Others friend and judge contributions based on the signature.
Frontend is copied from Wikipedia, hiding languages other than languages we read, for example, only Spanish and English. In addition to languages, the left column navigates to other coordinates, including discourses. The navigation method specifies the type of resource we are reading. For example, a selection of recent posts by friends is a social network, navigation by article titles is Wikipedia, by discourses – viewing alternative points of view on one topic. Other forms are simple sets of rules created by users and loaded into the fronted.
Navigation using orthogonal coordinates places closely related texts side by side. For example, all the top articles about something from different points of view will be placed next to each over in the discourse navigation panel. The proximity of closely related texts helps to fight copy-paste. Of the similar publications, the most popular will have a higher priority. It also allows a quick response to other people's publications.
Search engines and Web directories promote monopolies, including the monopolies of Google as a search engine and Wikipedia as a large directory. In the mid-90s, there were directories of the entire Internet, then search engines (with a hidden Web directory inside), but of the large catalogs, only Wikipedia remained, all other Wiki projects are many orders of magnitude smaller. W++ is potentially a very large system, a monopoly. W++ can absorb any contribution that can be parsed into W++, not only all Wikis and all smaller clones of W++. Users prefer the largest system, the largest catalog of everything that creates full-fledged search functionality inside W++.
The manifest can be partially implemented, but full implementation will be the most viable, i.e., capable of absorbing partial implementations of W++ and the rest of the Internet.
[1] Discourses are dominant points of view shared by a large number of people.
[2] The coordinate system described below roughly corresponds to a multidimensional OLAP cube The coordinates are poorly ordered, but generally have a graph structure formed by references usually directed along one dimension. The application of OLAP to social networks is not new. Novelty of W++ is the access of users to navigation through OLAP indexes.
This is an author translation of https://vignatovic.dreamwidth.org/444.html from Russian to English.
The same text is also published on Reddit: https://www.reddit.com/user/vignatovic/comments/umqsam/w_manifesto_a_single_engine_and_a_single_database/
Bitcoin address to support the author bc1qrccvvnjmm33y8u6axsw007739vdzsnzw88znx8
