GMJ project: the concept
ƒ RD AAOW FDL; 13.09.2023; 1:24
Содержание страницы:
The project concept
We’ve been accompanying the Grammar must joy community for quite some time. And at this point, the conclusion suggests itself that the result of its existence wasn’t only the opportunity to collect some conclusions about what the ideal joke or anecdote could be. The fact is that in the process of filling the community, we successfully created and tested a mechanism that still exists as a button in the uNot application. How does it work?
-
First, the VK community was created, which was filled manually from various sources. Text entries, usually not exceeding a thousand characters in length, were used as content. They were published either immediately or delayed (but this is not particularly important). During active seeding, participating in this community made sense; nowadays, publications in it occur much less frequently, which is why it is more of an auxiliary nature.
-
Then a channel appeared on Telegram, since relaying directly from VK was greatly complicated by its developers. In addition, this “distillation” made it possible to carry out high-quality filtration of the content. In the process, a template file appeared with the serial numbers of all “live” records, which is used in the application to speed up work (it excludes requests for non-existing records).
-
Finally, the GMJ query function has been added to the application. It allows you to periodically update the range of available entries, randomly select one of them and display it in the news log. In this case, the selected entry is crossed out from the list of available ones, which eliminates repetitions. The list is reset when the end of the range is reached. Therefore, the entire archive of posts rotates continuously, remaining fully accessible to the user.
In this form, this mechanism has been working properly for more than three years. It can potentially be applied to any content with certain critical characteristics:
- Text entries must be discrete, that is, not connected to each other logically or chronologically;
- Text entries must be autonomous: each of them must have independent meaning and practical value;
- Text entries must not exceed a thousand characters: This is a technical limitation of the mechanism in its current form.
- At least 80% of records should not completely lose relevance over time (i.e., they shouldn’t become useless).
As you understand, jokes easily fit these characteristics – in general, the concept was developed for them. But it is inapplicable to weather forecasts or quotes. We have already tried to implement this method for content of a different kind. But these attempts haven’t yet been successful